Baltimore’s StoCastic is using AI to take COVID-19 triage out of the ER and mobilize home visits

With the spread of COVID-19 in danger of pushing hospitals beyond their limits, healthcare systems are, among other things, looking to keep hospitals from becoming overcrowded — and keeping the people giving care from becoming overwhelmed.

Baltimore tech company StoCastic is bringing artificial intelligence to that effort at a healthcare system in Illinois: OSF Healthcare, a Peoria-based system which operates 14 healthcare facilities in the midwestern state, will use the company’s E-Triage tool to shift the process of designating the degree of a patient’s condition out of the emergency room.

The tool will allow nurses to triage patients over the phone, which means a patient can be at home when they report concerns about COVID-19. Using vital signs as well as machine learning that draws on electronic health record data and medical history, E-Triage can classify whether patients are at high-risk and need hospitalization, or are at lower risk and can remain at home for follow-up.

For the latter group, OSF Healthcare is also mobilizing pandemic healthcare workers who will go to patients’ homes in vans to further assess and get tests to patients. StoCastic’s E-CaseLink, a second tool that uses predictive modeling, will be used to help ensure the mobile units are being efficiently deployed. In this case, the workers will not enter the home, but rather drop off kits and conduct assessments via phone.

“It helps manage and optimize that workforce so that those pandemic healthcare workers can see as many patients and serve the largest population possible given their manpower,” said StoCastic cofounder Scott Levin, who is an associate professor of emergency medicine at the Johns Hopkins School of Medicine.

StoCastic was founded in 2016 by Levin and CEO Eric Hamrock, who previously worked together at the Johns Hopkins School of Medicine for over a decade. Receiving National Science Foundation funding through the Small Business Innovation Research program, the company is based out of Towson University Incubator and is also a member of ETC.

The company’s combination of using health record data and employing predictive analysis for healthcare teams is applied for other solutions to healthcare providers in other states, as well.

“Our goal is never to take the place of the clinician. The goal is to support the clinician,” Hamrock said.

With the spread of COVID-19, it was able to adapt those tools to a new use case at a critical time when the number of cases are rising toward a projected peak.

“We had the pieces of the puzzle,” Levin said. “Because we had the pieces, it was a lot faster for us to launch a good product.”